Background
Since the first case reported of the SARS-CoV-2 virus was reported in Mexico (February 27, 2020) to date (August 30th, 2020), there are 595,841 confirmed cases and 64,158 deaths [
1]. Nuevo Leon, a northern state comprising a population of 5,119,504 inhabitants and which includes one out of the three most populated metropolitan areas in Mexico, has registered 49,457 confirmed cases and 2,403 deaths (August 30th, 2020), placing it in the third position nationwide [
2]. Despite the increasing epidemiological surveillance efforts using the gold-standard quantitative polymerase chain reaction (PCR), testing is limited to symptomatic individuals, and it varies widely according to the test availability, even in the same country. Moreover, the occurrence of mild or asymptomatic infections leads to an underestimation of cases. Thus, seroepidemiological studies will allow better estimates of transmission dynamics, cumulative prevalence, and the proportion of the remaining susceptible populations [
3].
Seroprevalence of anti-SARS-CoV-2 antibodies is now available in several world regions. In Europe, firstly affected by SARS-CoV-2, Germany, Spain, and Switzerland determined a 1.22% (March–May 2020), 4.6% (May 2020), and 4.8% (April–May 2020) antibody prevalence, respectively [
4‐
6]. In the Americas, a 2.8% seroprevalence was reported in the state of California, United States, during April 2020, while 0.22% was observed in Brazil during May 2020 in a population-based survey [
7,
8]. However, these serological surveys have been conducted at different epidemic phases in each country, using several immunological assays, sample sizes, and study designs in selected populations (population-based, blood donors, health-care workers) [
5,
9,
10]. Thus, information derived from other regions is not a true reflection of the local epidemic's occurrence and transmission. To date, there is no epidemiological data regarding anti-SARS-CoV-2 prevalence in Mexico.
We conducted a cross-sectional serological survey in a population of government employees for the qualitative detection of IgM and IgG antibodies against SARS-CoV-2. We aimed to determine, as a primary endpoint, the prevalence of anti-SARS-CoV-2 antibodies and as a secondary endpoint, we defined the clinical and demographic characteristics associated with seroprevalence in the studied population. These results are now implemented for pandemic management in our regional context.
Discussion
Our findings from this regional cross-sectional seroprevalence study for SARS-CoV-2 indicate that the prevalence of IgM/IgG antibodies against the novel coronavirus is 5.9% in a city of the metropolitan area of Monterrey. This study is the first data available, since the first case registered in February 2020, on the prevalence of anti-SARS-CoV-2 antibodies in our country. Because our survey was designed to obtain information at a local level, regional differences in cultural habits and other factors might be observed against other areas in Mexico. As the third most populated state in the country, Monterrey's metropolitan area was considered a hotspot for epidemic transmission, and our findings may provide additional data to complete epidemiological surveillance.
Several studies have been published since April 2020 regarding the prevalence of anti-SARS-CoV-2 antibodies. Each one provided a snapshot that mirrors time and space-specific circumstances in the different studied populations and the associated characteristics. Our prevalence was higher than those reported in California, United States (April 2020), and Spain and Switzerland during last May 2020 [
5,
6], but lower than the 7.3% determined in the 1104 samples from the general population of Stockholm, Sweden at the end of April, three months after the pandemic arrival [
13]. Previous data for other regions or states in Mexico are not currently available. Given the asynchrony in COVID-19 transmission in our country, different seroprevalences might be found. Serosurveys are a useful tool to determine the previous exposure to SARS-CoV-2 and better estimate the true number of infections with and without symptoms. Particularly, Mexico has very limited PCR testing based on the sentinel epidemiological model used for epidemiological surveillance. This survey included a large sample, allowing determining an initial reference of previous exposure in asymptomatic individuals. Furthermore, sampling was made in a selected population that belongs to the exposed essential government workforce. Also, demographic characteristics such as age, sex, and city of residence of the sampled population were different from the overall population and may affect generalizability of the results. These characteristics will make it feasible for prospective and follow-up studies. In conjunction, the prevalence of 5.9%, implying that around 40,289 people were infected in Ciudad Guadalupe (population 682,880 in 2015) during July 2020 [
11], 10.3 times the average 3,898 PCR confirmed cases at the time of this survey (July 23, 2020) [
2]. Notably, Nuevo Leon has the higher rate of PCR test per million people in our country. Similar assumptions could be made for other regions as soon as local data become available.
In agreement with our findings, sex was not associated with anti-SARS-CoV-2 antibodies either in selected populations, health-care workers, and population-based studies. In contrast, age was associated with higher odds of having a positive test in Brazilian blood donors [
14], while the opposite was seen in Swiss children and adults > 65 years in a population-based survey [
6]. Consistently with our results, no association was observed in the professional category, previous contact with a COVID-19 case, and comorbidities [
15]. Although hypertension and type 2 diabetes were associated with a higher risk for poor outcomes in COVID-19 patients [
16], no relationship was found with positive IgM/IgG. Educational level, household size, and symptoms were associated with the presence of antibodies [
6,
14,
15]. In our study, as expected, previous COVID-19 symptoms were associated with seropositivity (OR 4.06, 95% CI 2.99–5.51,
P < 0.001), although a higher OR was reported by Garcia-Bateiro et al. (OR 8.8, 95% CI 4.41–17.73,
P < 0.0001) [
15]. Among the symptoms reported, anosmia showed the strongest association in our analysis. Asymptomatic cases (59.1%) compared with other countries were higher than the 32.7%, and 52.4% reported in the general population and health-care workers in Spain, respectively [
5,
15]. As much as approximately 80% of infections are expected to be asymptomatic or mild [
17].
Recent studies indicate some limitations of antibody testing, besides the inherent diagnostic performance. Although antibodies are expected to correlate with infection and disease presentation, a recent Chinese study showed that some infected individuals did not produce neutralizing humoral evidence of SARS-CoV-2 infection. In 10 (5.7%) out of 175 COVID-19 recovered patients who experienced mild symptoms, neutralizing antibodies against SARS-CoV-2 were unable to cross-react with the SARS-CoV-2 virus [
18]. Thus, other immune responses, such as T cells and cytokines, may be involved in recovery. Furthermore, Long e
t al. reported than 40% of asymptomatic patients became seronegative for IgG in the early convalescent phase [
19] suggesting a weaker humoral response that might have implications in the results of serological surveys confounding the true exposure rate.
Our serosurvey has some limitations that may be considered. First, we performed an isolated IgM/IgG test, obtaining a cross-sectional seroprevalence. A prospective and longitudinal survey will be highly valuable in following seroconversion and determining household transmission and lockdown effect in the long-term. Second, children were not included in the survey; additional studies will be needed to determine SARS-CoV-2 spread in this group. Finally, due to the study design, these results represent our sampled population's prevalence estimates and do not necessarily reflect the entire state prevalence. Otherwise, as strengths for the study, we used an appropriate and powered sample size to provide a reference prevalence value. Also, it seems we surveyed at the right epidemic moment (WHO Phase 3). Besides, we used a high-sensitivity chemiluminescent immunoassay, as reported by Muench et al. [
20], which has shown a better diagnostic performance than other tests.
Conclusions
In conclusion, this study provides a regional estimation of SARS-CoV-2 dissemination in the densely populated northern region of our country. Six in ten infections were asymptomatic, and despite the significant impact of COVID-19 in Mexico, seroprevalence is still very low and, more importantly, far from herd immunity. Epidemiological surveillance and preventive measures to avoid transmission, such as social distancing and facial masks use, should still be enforced in the population.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.